Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

VaSST claims that symbolic regression can be solved by variational inference over continuously relaxed expression trees, recovering true equations from noisy data with quantified uncertainty.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

VaSST uses variational inference over continuously relaxed symbolic trees to recover closed-form expressions from noisy data, reporting competitive structural recovery and predictive accuracy on simulated and Feynman benchmarks.

T0 review reviewed 2026-08-02 challenge →

load-bearing objection Solid variational SR machinery, but the 'exact recovery' headline fails on the paper's own appendix; worth refereeing, not worth citing yet. the 3 major comments →

arxiv 2602.23561 v2 pith:ZY2P5Y3W submitted 2026-02-27 stat.ME cs.LGcs.SCstat.COstat.ML

VaSST: Variational Inference for Symbolic Regression using Soft Symbolic Trees

Somjit Roy , Pritam Dey , Bani K. Mallick This is my paper
classification stat.ME cs.LGcs.SCstat.COstat.ML MSC 62F1562J05
keywords symbolic regressionvariational inferencesoft symbolic treesuncertainty quantificationcontinuous relaxationBayesian inferencescientific discoveryparsimony
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

In symbolic regression, the goal is to recover a closed-form equation from data, but the space of expressions is combinatorial and existing probabilistic methods are either heuristic or slow. VaSST claims to solve this by treating each node of a symbolic tree as a soft probability distribution over operators and features, which turns the discrete search into a smooth optimization problem inside a variational inference framework. The method's load-bearing assertion is empirical: across two simulated laws and four Feynman equations, at noiseless and noisy levels, the minimum-RMSE model from VaSST recovers the correct symbolic expression while also producing a ranked list of plausible alternative structures. If this holds, it would give scientists a practical, uncertainty-aware tool for equation discovery that is both faster and more parsimonious than current Bayesian approaches.

Core claim

The paper's central claim is that symbolic regression can be recast as a scalable variational inference problem. VaSST replaces discrete expression trees with soft symbolic trees, in which each node's operator and feature choices are continuous probability vectors (via Binary Concrete and Gumbel-Softmax relaxations). The variational objective—a stochastic approximation of the ELBO that evaluates the marginal likelihood at the soft design matrix—is optimized by gradient descent under a temperature annealing schedule. After optimization, hard trees are sampled from the learned variational distribution and ranked by minimum in-sample RMSE. On two simulated laws and four equations from the Feynm

What carries the argument

Soft symbolic trees: a continuous relaxation of a symbolic expression tree in which each node carries a soft one-hot distribution over expansion indicators (Binary Concrete), operators, and input features (Gumbel-Softmax). The tree is evaluated with a soft-gating equation that mixes leaf features and unary/binary operations; this makes the entire design matrix differentiable in the variational parameters. The depth-dependent Bernoulli prior p_ζ = α(1+d_ζ)^{-δ} provides the parsimony mechanism.

Load-bearing premise

The method maximizes a relaxed ELBO that evaluates the marginal likelihood at soft symbolic trees; the paper gives no bound on how far this surrogate optimum is from the true posterior over hard trees, so the structural recovery and uncertainty claims rest on the relaxation gap being small.

What would settle it

A direct check is to enumerate all hard trees for a small problem (depth 2–3, few features), compute the exact ELBO at the optimized variational parameters, and compare it to the relaxed ELBO value and to the true posterior mode. If the relaxed optimum is far in function value or in structure from the true posterior, the UQ claim fails. Additionally, test VaSST on a low-noise dataset generated by an expression with nested divisions or exponentiation that the soft relaxation may represent poorly; failure to recover the true expression would undercut the recovery claim.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Symbolic regression becomes a gradient-based optimization problem rather than a discrete combinatorial search, scaling to larger datasets and higher-dimensional feature spaces.
  • The method yields a posterior distribution over symbolic structures, enabling uncertainty quantification in equation discovery instead of a single point estimate.
  • The depth-dependent split prior acts as a built-in parsimony regularizer, favoring simpler expressions and reducing overfitting.
  • On the tested benchmarks, VaSST recovers exact ground-truth equations across all noise levels, whereas many competing methods produce overly complex or incorrect forms.
  • VaSST is substantially faster than existing Bayesian symbolic regression methods (BMS and BSR), making fully probabilistic SR practical on moderately sized datasets.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The relaxed ELBO could systematically favor smooth blends of operators; a testable consequence is that VaSST's structural recovery may degrade on equations requiring sharp discontinuities or nested compositions that the soft-gating evaluation at temperature 0.5 cannot approximate.
  • Ranking candidate expressions by minimum in-sample RMSE is a pragmatic selection rule but may overfit to noise; an alternative would weight structures by posterior evidence rather than predictive fit, which the paper does not explore.
  • The depth-dependent prior is the main parsimony lever, so a sensitivity analysis on α and δ would reveal how much of the recovery success is due to prior regularization versus the variational objective itself.
  • Because hard trees are sampled after optimization, the framework could be extended to full posterior predictive distributions and Bayesian model averaging over expressions, a direction the paper leaves implicit.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes VaSST, a variational-inference framework for symbolic regression in which discrete symbolic trees are relaxed into 'soft symbolic trees' via Binary Concrete and Gumbel-Softmax reparameterizations. The authors derive an ELBO consisting of an analytically intractable marginal-likelihood term approximated by Monte Carlo and a fully analytic KL divergence, and optimize it with black-box variational inference. They claim that the method scales better than MCMC-based Bayesian SR, provides uncertainty quantification over symbolic structures, and, in Section 5.2, that 'across all equation datasets and noise levels, the minimum RMSE model produced by VaSST recovers the correct symbolic expression.' The paper evaluates VaSST on two simulated equations and four Feynman equations against BMS, BSR, QLattice, gplearn, and DEAP.

Significance. If the central claims were fully supported, the contribution would be significant: a scalable, fully probabilistic SR method with a principled prior over tree depth, an analytic KL divergence, and a coherent mechanism for sampling hard symbolic trees for uncertainty quantification. The authors ship a derivation of the marginal likelihood and KL terms that is standard and appears correct, and the reported predictive RMSEs are competitive while runtimes are substantially lower than the Bayesian competitors. The conceptual move of relaxing discrete symbolic structure for variational SR is interesting and worth publishing. However, the headline claim of exact symbolic recovery is contradicted by the paper's own appendix, and the variational surrogate is not shown to be faithful to the discrete posterior; these issues are load-bearing for both the recovery and the uncertainty-quantification claims.

major comments (3)
  1. [§5.2, Table 3, Appendix K] The claim that 'across all equation datasets and noise levels, the minimum RMSE model produced by VaSST recovers the correct symbolic expression' is contradicted by the expressions reported in Appendix K. For FCE under the noiseless setting, Table K.7 gives VaSST's min-RMSE expression as −0.00285 + 0.00318 sin(2Bv/E_f) + 0.0012θ + 1.068(q sin(θ)Bv + E_f q). This is not q(E_f + vB sinθ): the coefficient on the main term is 1.068 rather than 1, and there are spurious sin(2Bv/E_f) and θ terms. The same non-exact expression is the top-1 sample in Table K.16. For CPE noiseless, the top-1 expression in Table K.15 contains a spurious G²m₁³m₂r₁⁻² term and coefficients −0.997 and 1.011; for FTC noiseless, the top-1 expression in Table K.17 has coefficients 0.999 and −0.905 plus an intercept and omits the exact coefficient 1 on κAT₂/d. Hence the ✓ entries for VaSST in Table 3 are not supported by
  2. [§4, Eq. (14)] The stochastic ELBO replaces the hard design matrix T in log p(y|T) with the soft design matrix T_soft obtained from Binary Concrete/Gumbel-Softmax draws, but no bound or quantitative justification is given for treating this surrogate as faithful to the discrete ELBO in Eq. (9). The optimization is performed at temperatures annealed only to τ=0.5, so the final objective is still evaluated at continuous, mixed trees rather than hard symbolic trees. Unless the relaxation gap is shown to be small at the final temperature, the optimized variational parameters φ* need not represent a posterior over discrete symbolic structures, and the subsequent sampling of hard trees in Algorithm F.5 does not inherit a valid posterior interpretation. This affects both the structural-recovery claims and the uncertainty-quantification claims. The authors should either provide a theoretical bound or an empiric
  3. [§4, Algorithm F.5 and §5.2] The reported 'minimum RMSE model' is selected post hoc as the best of H=2000 hard-tree samples ranked by in-sample RMSE. This is a selection procedure, not a posterior summary, and it can substantially inflate apparent recovery: among 2000 draws, the single best-fitting structure may match the data by chance even if the posterior mass on the true expression is negligible. The paper should report the posterior probability or frequency of the true symbolic expression among the H samples, or use Bayesian model averaging / posterior-weighted predictive summaries, before claiming that VaSST 'recovers' the correct expression. Without this, the central recovery claim is not a statement about the inferred posterior.
minor comments (5)
  1. [§3.2 / Eq. (3)] The term 'unary operator pruning' in the deterministic pruning step is described only in words; a formal definition of the pruning map p would improve reproducibility, especially because the right subtree is removed when o_jζ is unary.
  2. [Appendix G.2] The cross-reference 'last two expressions in (4)' should refer to the KL derivations in Appendix G.2, not to Eq. (4); the wording is confusing.
  3. [Section 5.1] The text says VaSST 'accurately recovers the true structure' for Eq. (15), but the reported coefficients in Table 1 and Appendix I.1 are not the true coefficients (e.g., 0.61 vs. 0.5 in Table 1). If the claim is structural recovery up to coefficient estimation, this should be stated explicitly and a tolerance or coefficient-error metric should be supplied.
  4. [Appendix I.3 / K.4] Several top-5 expressions contain duplicate terms (e.g., Table I.3: '+0.001381 sin(x1) + 0.001381 sin(x1)') and unusual factorizations (e.g., Table K.17, rank 2). While not incorrect, the lack of simplification makes reported expressions harder to audit; a canonical simplification step would help.
  5. [Throughout] There are minor typos: 'Dension et al., 1998' should be 'Denison et al., 1998'; 'constat' in Table I.3; 'BMSfailed' in Table K.10; and the abstract says 'superior performance' while the conclusion says 'strong structural recovery'. These should be corrected.

Circularity Check

0 steps flagged

No circular derivation: the variational objective and hard-tree sampling are constructed from data, priors, and standard relaxations, not from target equations; the exact-recovery overclaim is an internal validation inconsistency, not a circular reduction.

full rationale

The derivation chain is self-contained against the data and fixed priors. The model in (1)-(2) is a linear ensemble of symbolic trees with a conjugate Normal-Inverse-Gamma prior; the marginal likelihood (6)/(G.3) is a standard closed-form integration, with no appearance of the target expressions. The variational ELBO (9) is decomposed into analytic KL terms (10)-(11) plus an intractable expected log-marginal-likelihood term, which is then approximated in (14) using Binary Concrete/Gumbel-Softmax relaxations. The relaxed objective is a surrogate for the true ELBO, and the paper does not bound the relaxation gap, but this is an approximation-error/statistical-validity concern, not circularity: the optimum of (14) is not equal to the true posterior or to the target equation by construction. Hard symbolic trees are sampled from the optimized variational parameters and ranked by minimum in-sample RMSE (Section 4, Algorithm F.5). This is post-hoc model selection rather than posterior-weighted prediction, and it can inflate apparent recovery, but the selected expression is data-dependent and is not definitionally identical to the target. The paper's self-citations are standard Bayesian references used for priors and are not load-bearing uniqueness claims; no ansatz is smuggled in via self-citation, and no known result is merely renamed. The strongest concern is factual: Section 5.2 claims that 'the minimum RMSE model produced by VaSST recovers the correct symbolic expression,' yet the paper's own Appendix K.4 shows recovered FCE expressions such as '-0.00285+0.00318 sin(2Bv/E_f)+0.0012θ+1.068(q sinθ Bv + E_f q)', which are not equal to q(E_f + vB sinθ). This is an internal validation inconsistency and an overclaim, not circular reasoning. Score 0 for circularity.

Axiom & Free-Parameter Ledger

3 free parameters · 4 axioms · 1 invented entities

The paper's central claim rests on a conjugate marginalization (standard math), a mean-field approximation, and a relaxation-faithfulness assumption. The latter is the most fragile: no bound is shown for the gap between the relaxed and true ELBO. The depth prior hyperparameters and capacity choices (K,D) are fixed by hand, not fitted to the target expressions.

free parameters (3)
  • depth prior hyperparameters α, δ = 0.95, 2.00
    Control the depth-adaptive split probability p_ζ = α(1+d_ζ)^{-δ} in Eq. (3); chosen by hand and load-bearing for structural parsimony.
  • ensemble size K and max depth D = K=3, D=3
    Model capacity used for all experiments; chosen by hand in Appendix H.1.
  • temperature annealing schedule = τ_start=1.0, τ_end=0.5, T_τ=1500
    Annealing of Binary Concrete/Gumbel-Softmax temperatures in Eq. (12); chosen by hand; affects the relaxation bias.
axioms (4)
  • standard math Conjugate Normal-Inverse-Gamma marginalization yields the marginal likelihood formula (G.3).
    Relies on standard Gaussian-Gamma conjugacy; derivation in Appendix G.1.
  • domain assumption The mean-field variational family (8) is a tractable and sufficiently flexible approximation to the true posterior over symbolic structures.
    The posterior over trees is highly multimodal; mean-field assumes independence across nodes, an approximation that may weaken UQ.
  • ad hoc to paper The Binary Concrete/Gumbel-Softmax relaxation in Eq. (12) yields a soft design matrix whose ELBO (14) is a faithful surrogate for the discrete ELBO.
    No theoretical guarantee or bias bound is provided; this is the central methodological assumption.
  • domain assumption The operator set O and maximum depth D are sufficient to represent the target expressions.
    If the true law requires an operator not in O or depth beyond D, recovery is impossible; validated only on four simple Feynman equations.
invented entities (1)
  • Soft symbolic tree (S_soft) independent evidence
    purpose: Continuous relaxation of symbolic expression trees enabling gradient-based variational inference.
    It is the paper's core methodological construct, empirically validated on benchmarks; not an unexplained physical entity.

reviewed 2026-08-02 · how reviews work

0 comments
Cite this review

Pith. "Pith review of VaSST: Variational Inference for Symbolic Regression using Soft Symbolic Trees." pith.science (2026). https://pith.science/paper/ZY2P5Y3W

@misc{pith2026260223561,
  author       = {Pith},
  title        = {Pith review of: VaSST: Variational Inference for Symbolic Regression using Soft Symbolic Trees},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZY2P5Y3W}},
  note         = {Machine review of arXiv:2602.23561}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Symbolic regression (SR) has gained recent traction in AI-driven scientific discovery for learning closed-form physical laws. Yet existing methods are dominated by heuristic search or data-intensive approaches that often assume low-noise regimes and lack principled uncertainty quantification, while fully probabilistic SR formulations remain scarce. We introduce a scalable probabilistic framework for SR, VaSST, based on variational inference. VaSST uses soft symbolic trees, a continuous relaxation of symbolic expression trees in which discrete operator and feature assignments are replaced by probability distributions over allowable components. This transforms combinatorial symbolic search through an astronomically large expression space into efficient gradient-based optimization while preserving a coherent probabilistic interpretation. The learned soft representations induce posterior distributions over symbolic structures, enabling uncertainty quantification across plausible symbolic forms through posterior-aware symbolic model selection. On simulated experiments and the Feynman Symbolic Regression Database, VaSST achieves strong structural recovery and predictive accuracy compared to state-of-the-art competing SR methods.

Figures

Figures reproduced from arXiv: 2602.23561 by Bani K. Mallick, Pritam Dey, Somjit Roy.

Figure 1
Figure 1. Figure 1: Symbolic tree representation. Now, for E ∈ E and an instance x ∈ R p of the features in X , we define g(x;T (E)) to be the real-valued evaluation of the symbolic expression T (E) at x. With this setup in place, we formally present the VaSST modeling framework. 3 The VaSST Model We outline the VaSST model which comprises two major components: (i) a symbolic ensemble consisting of a linear regression of the … view at source ↗
Figure 2
Figure 2. Figure 2: A live representation of the prior over S (2) j and the induced prior over T (Ej). In S (2) j , the children nodes ζ = 3,4 of ζ = 1 are pruned using p as e j1 = 0 to obtain the valid symbolic tree T (Ej). Valid symbolic tree via deterministic pruning. The skeleton S (D) j is an ambient representation. We obtain the corresponding valid symbolic tree T (Ej) by applying a deterministic pruning operator p whic… view at source ↗
Figure 3
Figure 3. Figure 3: Computational scalability of VaSST, BMS, BSR. VaSST effectively balances structural learning and predictive accuracy, achieving consistently low out-of-sample RMSEs across all noise levels when learning (15) ( [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Are you sure? A Comprehensive and Comprehensible Survey of Uncertainty Quantification in Symbolic Regression

    cs.LG 2026-06 unverdicted novelty 6.0

    A survey claiming to be the first comprehensive review of uncertainty quantification in symbolic regression, organized into three research directions and noting the area remains underexplored.

Reference graph

Works this paper leans on

45 extracted references · 5 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Mixing time of exponential random graphs

    Shankar Bhamidi, Guy Bresler, and Allan Sly. Mixing time of exponential random graphs. In 49th Annual IEEE Symposium on Foundations of Computer Science, pages 803--812, 2008

    2008

  2. [2]

    Blei, Andrew Y

    David M. Blei, Andrew Y. Ng, and Michael I. Jordan. Latent dirichlet allocation . Journal of Machine Learning Research, 3: 0 993--1022, 2003

    2003

  3. [3]

    Blei, Alp Kucukelbir, and Jon D

    David M. Blei, Alp Kucukelbir, and Jon D. McAuliffe. Variational Inference: A Review for Statisticians . Journal of the American Statistical Association, 112 0 (518): 0 859--877, 2017

    2017

  4. [4]

    Deep learning methods for protein function prediction

    Frimpong Boadu, Ahhyun Lee, and Jianlin Cheng. Deep learning methods for protein function prediction . PROTEOMICS, 25 0 (1-2): 0 2300471, 2025

    2025

  5. [5]

    Bagging predictors

    Leo Breiman. Bagging predictors. Machine Learning, 24 0 (2): 0 123--140, 1996

    1996

  6. [6]

    Leo Breiman, Jerome Friedman, R. A. Olshen, and Charles J. Stone. Classification and Regression Trees (1st ed.) . Chapman and Hall/CRC, 1984

    1984

  7. [7]

    An approach to symbolic regression using feyn

    Kevin Brol s, Meera Vieira Ren \'e , Machado, Chris Cave, Jaan Kasak, Valdemar Stentoft-Hansen, Victor Galindo Batanero, Tom Jelen, and Casper Wilstrup. An approach to symbolic regression using feyn. arXiv:2104.05417, 2021

    Pith/arXiv arXiv 2021

  8. [8]

    Brunton, Joshua L

    Steven L. Brunton, Joshua L. Proctor, and J. Nathan Kutz. Discovering governing equations from data by sparse identification of nonlinear dynamical systems. Proceedings of the National Academy of Sciences, 113 0 (15): 0 3932--3937, 2016

    2016

  9. [9]

    Machine learning for molecular and materials science

    Keith T Butler, Daniel W Davies, Hugh Cartwright, Olexandr Isayev, and Aron Walsh. Machine learning for molecular and materials science. Nature, 559 0 (7715): 0 547--555, 2018

    2018

  10. [10]

    Chipman, Edward I

    Hugh A. Chipman, Edward I. George, and Robert E. McCulloch. Bayesian CART Model Search . Journal of the American Statistical Association, 93 0 (443): 0 935--948, 1998

    1998

  11. [11]

    Chipman, Edward I

    Hugh A. Chipman, Edward I. George, and Robert E. McCulloch. BART: Bayesian additive regression trees . The Annals of Applied Statistics, 4 0 (1): 0 266--298, 2010

    2010

  12. [12]

    J. W. Davidson, D. A. Savic, and G. A. Walters. Symbolic and numerical regression: experiments and applications. Information Sciences, 150 0 (1–2): 0 95--117, 2003

    2003

  13. [13]

    David G. T. Dension, Bani K. Mallick, and Adrian F. M. Smith. A Bayesian CART algorithm. Biometrika, 85 0 (2): 0 363--377, 1998

    1998

  14. [14]

    Feynman, R.B

    R.P. Feynman, R.B. Leighton, and M. Sands. The Feynman Lectures on Physics, Vol. I: The New Millennium Edition: Mainly Mechanics, Radiation, and Heat. Number v. 1. Basic Books, 2015

    2015

  15. [15]

    DEAP : Evolutionary algorithms made easy

    F \'e lix-Antoine Fortin, Fran c ois-Michel De Rainville, Marc-Andr \'e Gardner, Marc Parizeau, and Christian Gagn \'e . DEAP : Evolutionary algorithms made easy. Journal of Machine Learning Research, 13 0 (70): 0 2171--2175, 2012

    2012

  16. [16]

    Black box variational inference with a deterministic objective: Faster, more accurate, and even more black box

    Ryan Giordano, Martin Ingram, and Tamara Broderick. Black box variational inference with a deterministic objective: Faster, more accurate, and even more black box . Journal of Machine Learning Research, 25 0 (18): 0 1--39, 2024

    2024

  17. [17]

    Practical Variational Inference for Neural Networks

    Alex Graves. Practical Variational Inference for Neural Networks . In Advances in Neural Information Processing Systems, volume 24, 2011

    2011

  18. [18]

    A bayesian machine scientist to aid in the solution of challenging scientific problems

    Roger Guimer \`a , Ignasi Reichardt, Antoni Aguilar-Mogas, Francesco A Massucci, Manuel Miranda, Jordi Pallar \`e s, and Marta Sales-Pardo. A bayesian machine scientist to aid in the solution of challenging scientific problems. Science advances, 6 0 (5): 0 eaav6971, 2020

    2020

  19. [19]

    Categorical reparameterization with gumbel-softmax

    Eric Jang, Shixiang Gu, and Ben Poole. Categorical reparameterization with gumbel-softmax. arXiv:1611.01144, 2016

    Pith/arXiv arXiv 2016

  20. [20]

    Jefferys and James O

    William H. Jefferys and James O. Berger. Ockham's Razor and Bayesian Analysis . American Scientist, 80 0 (1): 0 64--72, 1992. ISSN 00030996

    1992

  21. [21]

    Bayesian Symbolic Regression

    Ying Jin, Weilin Fu, Jian Kang, Jiadong Guo, and Jian Guo. Bayesian Symbolic Regression . arXiv:1910.08892, 2020

    Pith/arXiv arXiv 1910

  22. [22]

    An introduction to variational methods for graphical models

    Michael I Jordan, Zoubin Ghahramani, Tommi S Jaakkola, and Lawrence K Saul. An introduction to variational methods for graphical models . Machine learning, 37 0 (2): 0 183--233, 1999

    1999

  23. [23]

    End-to-end symbolic regression with transformers

    Pierre-Alexandre Kamienny, St \'e phane d'Ascoli, Guillaume Lample, and Fran c ois Charton. End-to-end symbolic regression with transformers. In Advances in Neural Information Processing Systems, 2022

    2022

  24. [24]

    Michael F. Korns. Accuracy in symbolic regression. In Genetic Programming Theory and Practice IX, pages 129--151. 2011

    2011

  25. [25]

    Genetic programming as a means for programming computers by natural selection

    John R Koza. Genetic programming as a means for programming computers by natural selection. Statistics and computing, 4 0 (2): 0 87--112, 1994

    1994

  26. [26]

    Automatic differentiation variational inference

    Alp Kucukelbir, Dustin Tran, Rajesh Ranganath, Andrew Gelman, and David M Blei. Automatic differentiation variational inference . Journal of machine learning research, 18 0 (14): 0 1--45, 2017

    2017

  27. [27]

    MCMC for multi-modal distributions

    Krzysztof atuszy \'n ski, Matthew T Moores, and Timoth \'e e Stumpf-F \'e tizon. MCMC for multi-modal distributions. arXiv:2501.05908, 2025

    Pith/arXiv arXiv 2025

  28. [28]

    DARTS : Differentiable architecture search

    Hanxiao Liu, Karen Simonyan, and Yiming Yang. DARTS : Differentiable architecture search. In International Conference on Learning Representations, 2019

    2019

  29. [29]

    Decoupled Weight Decay Regularization

    Ilya Loshchilov and Frank Hutter. Decoupled Weight Decay Regularization . In International Conference on Learning Representations, 2019

    2019

  30. [30]

    Maddison, Andriy Mnih, and Yee Whye Teh

    Chris J. Maddison, Andriy Mnih, and Yee Whye Teh. The concrete distribution: A continuous relaxation of discrete random variables. arXiv:1611.00712, 2017

    Pith/arXiv arXiv 2017

  31. [31]

    Mundhenk, Claudio Prata Santiago, Soo Kyung Kim, and Joanne Taery Kim

    Brenden K Petersen, Mikel Landajuela Larma, Terrell N. Mundhenk, Claudio Prata Santiago, Soo Kyung Kim, and Joanne Taery Kim. Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients. In International Conference on Learning Representations, 2021

    2021

  32. [32]

    Raissi, P

    M. Raissi, P. Perdikaris, and G.E. Karniadakis. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378: 0 686--707, 2019

    2019

  33. [33]

    Automatic differentiation: Techniques and applications

    Louis B Rall. Automatic differentiation: Techniques and applications. Springer, 1981

    1981

  34. [34]

    Rajesh Ranganath, Sean Gerrish, and David M. Blei. Black Box Variational Inference . In International Conference on Artificial Intelligence and Statistics, volume 33, pages 814--822, 2014

    2014

  35. [35]

    Carl Edward Rasmussen and Christopher K. I. Williams. Gaussian Processes for Machine Learning. MIT Press, 2006

    2006

  36. [36]

    VaRT : Variational regression trees

    Sebastian Salazar. VaRT : Variational regression trees. In Advances in Neural Information Processing Systems, 2023

    2023

  37. [37]

    Distilling free-form natural laws from experimental data

    Michael Schmidt and Hod Lipson. Distilling free-form natural laws from experimental data. Science, 324 0 (5923): 0 81--85, 2009

    2009

  38. [38]

    gplearn : Genetic Programming in Python , 2016

    Trevor Stephens. gplearn : Genetic Programming in Python , 2016. URL https://github.com/trevorstephens/gplearn

    2016

  39. [39]

    Regression Shrinkage and Selection via the Lasso

    Robert Tibshirani. Regression Shrinkage and Selection via the Lasso . Journal of the Royal Statistical Society. Series B (Methodological), 58 0 (1): 0 267--288, 1996

    1996

  40. [40]

    AI Feynman: A physics-inspired method for symbolic regression

    Silviu-Marian Udrescu and Max Tegmark. AI Feynman: A physics-inspired method for symbolic regression . Science Advances, 6 0 (16): 0 eaay2631, 2020

    2020

  41. [41]

    Marco Virgolin and Solon P. Pissis. Symbolic regression is np-hard. In Proceedings of the Genetic and Evolutionary Computation Conference, 2022

    2022

  42. [42]

    Wainwright and Michael I

    Martin J. Wainwright and Michael I. Jordan. Graphical Models, Exponential Families, and Variational Inference . Foundations and Trends in Machine Learning, 1 0 (1–2): 0 1--305, 2008

    2008

  43. [43]

    Exploring the mathematic equations behind the materials science data using interpretable symbolic regression

    Guanjie Wang, Erpeng Wang, Zefeng Li, Jian Zhou, and Zhimei Sun. Exploring the mathematic equations behind the materials science data using interpretable symbolic regression. Interdisciplinary Materials, 3 0 (5): 0 637--657, 2024

    2024

  44. [44]

    Simple statistical gradient-following algorithms for connectionist reinforcement learning

    Ronald J Williams. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning, 8 0 (3): 0 229--256, 1992

    1992

  45. [45]

    Machine Learning Methods for Weather Forecasting: A Survey

    Huijun Zhang, Yaxin Liu, Chongyu Zhang, and Ningyun Li. Machine Learning Methods for Weather Forecasting: A Survey . Atmosphere, 16 0 (1), 2025

    2025

This paper was first reviewed by deepseek-v4-flash on August 2, 2026.